Will an AI ever be able to make correct moral decisions? Or is that something only a human can do?